DRIFT: Data Reduction via Informative Feature Transformation- Generalization Begins Before Deep Learning starts

Modern deep learning architectures excel at optimization, but only after the data has entered the network. The true bottleneck lies in preparing the right input: minimal, salient, and structured in a way that reflects the essential patterns of the data. We propose DRIFT (Data Reduction via Informative Feature Transformation), a novel preprocessing technique inspired by vibrational analysis in physical systems, to identify and extract the most resonant modes of input data prior to training. Unlike traditional models that attempt to learn amidst both signal and noise, DRIFT mimics physics perception by emphasizing informative features while discarding irrelevant elements. The result is a more compact and interpretable representation that enhances training stability and generalization performance. In DRIFT, images are projected onto a low-dimensional basis formed by spatial vibration mode shapes of plates, offering a physically grounded feature set. This enables neural networks to operate with drastically fewer input dimensions (~ 50 features on MNIST and less than 100 on CIFAR100) while achieving competitive classification accuracy. Extensive experiments across MNIST and CIFAR100 demonstrate DRIFT's superiority over standard pixel-based models and PCA in terms of training stability, resistance to overfitting, and generalization robustness. Notably, DRIFT displays minimal sensitivity to changes in batch size, network architecture, and image resolution, further establishing it as a resilient and efficient data representation strategy. This work shifts the focus from architecture engineering to input curation and underscores the power of physics-driven data transformations in advancing deep learning performance.

Architect Your Landscape Approach (AYLA) for Optimizations in Deep Learning

Stochastic Gradient Descent (SGD) and its variants, such as ADAM, are foundational to deep learning optimization, adjusting model parameters using fixed or adaptive learning rates based on loss function gradients. However, these methods often face challenges in balancing adaptability and efficiency in non-convex, high-dimensional settings. This paper introduces AYLA, a novel optimization technique that enhances training dynamics through loss function transformations. By applying a tunable power-law transformation, AYLA preserves critical points while scaling loss values to amplify gradient sensitivity, accelerating convergence. We further propose a dynamic (effective) learning rate that adapts to the transformed loss, improving optimization efficiency. Empirical tests on finding minimum of a synthetic non-convex polynomial, a non-convex curve-fitting dataset, and digit classification (MNIST) demonstrate that AYLA surpasses SGD and ADAM in convergence speed and stability. This approach redefines the loss landscape for better optimization outcomes, offering a promising advancement for deep neural networks and can be applied to any optimization method and potentially improve the performance of it.